How Does Anti-Spam Software Do Its Work?
Over the past 15 years or so, the problem with overwhelming spam—countless unwanted and unrequested emails—has lessened for the most part. Before anti-spam software became commonplace, spammers could flood our email inboxes with their messages. Now, the wave of spam has been slowed down. Thank you, anti-spam software.
Here's a look at how a coordinated anti-spam effort works.
ISPs and spam.
To begin with, anti-spam protection for an everyday computer such as yours starts with your Internet Service Provider—an ISP such as Cox Cable, AT&T, etc. They use sophisticated software on their email servers to catch spam as it comes in, and they try to prevent it from ever reaching you.
Most ISPs that offer email accounts include anti-spam software for their users. But since that's not their primary business (getting you connected is), it's not as robust as you need it to be. That's why it's important to take matters into your own hands to make sure you're protected.
That's where individual anti-spam software comes into play, which runs right from your email program, whether that's Gmail, Outlook, or some other program. It flags emails that it considers to be suspicious content and sends it straight into a spam folder, not your regular inbox...setting it aside for you to investigate later.
One approach developed long ago is called content filtering, where anti-spam software examines the body and subject line (headline) of an email, as well as words in the message.
The subject line might say "Dentist Appointment Reminder" or "Win a Trip to Hawaii!" (Guess which one might be spam.) Content filtering examines the headline against an extensive internal database of words and terms that spammers use, including obvious ones such as:
- Extra Cash
- Viagra formula
- Rolex discounts
Content filters only block emails that contain flagged, "spammy" words, word combinations and word phrases. So spammers will alter word spellings and use other tricks to avoid getting blocked out so they can sneak their messages through to you.
Anti-spam companies upped their game long ago with a Bayesian spam filter, a highly complex and more sophisticated approach to catching spam. It's named after an English statistician from the 1700s named Thomas Bayes and it's based on probability theory and statistics.
But all you really need to know is that by analyzing an email, the Bayesian filter calculates the probability of that message being spam and grades its "spamicity"—if its spamicity goes over a certain threshold, that word and message are flagged as spam.
It gets even better. Over time, the Bayesian spam filter builds a custom database, by learning the select words that email customers consider to be spam and which they don't.
This means spam war!
Marketers, who aren't necessarily spammers but can be annoying emailers, are constantly learning what words are being treated as spam and try to avoid them. In all fairness, not all marketing messages are actual spam—they may just look like spam or, more accurately, sound like spam by anti-spam software.
There are special websites that keep marketers informed on the most-current "spammy" words and phrases to avoid using in online promotions. Here's a sample list:
- 100% free
- Act now!
- ALL CAPITALS (use of)
- All natural
- As seen on
- Bad credit
The list came with this advice they give to their readers:
"We have found that Server filters like Spam Assassin, Postini; Gateway filters like IronPort, Barracuda; along with Desktop Client Filters like Outlook watch out for these phrases. These words and phrases are more likely to trigger a spam score and cause email blocking or bulking issues."
Our advice to you? Keep your anti-spam software up to date.